For teams who carry their own pager
Find out it broke before your customers tell you
Checks that notice, an on-call rota that wakes the right person, and the page customers read while you fix it. One tool instead of three — and on-call is free, with no per-seat bill.
No credit card. Live in a few minutes.
What’s free forever
- ✓On-call rotas with layers — weekdays one person, nights another
- ✓Escalation that keeps going until somebody acknowledges
- ✓Unlimited people, and no per-seat charge, ever
- ✓Slack, Telegram, webhooks and browser push — uncounted
- ✓10 checks: a URL, a port, a certificate, or a job that calls in
- ✓Route by severity: P1 wakes someone, P4 waits until morning
- ✓Inbound webhooks from Sentry, Grafana, CI and your own scripts
- ✓A status page on your own domain, badge and 90-day uptime
How it works
Step 1
Say what you have
Name the parts customers would notice — API, Web, Billing — not every microservice. Everything else hangs off those names.
Step 2
Say who to reach, and when
Put people in a rota, order the escalation, and decide which severity goes where. Everyone confirms their own phone or Telegram, so nothing is sent to an address nobody checked.
Step 3
Let it run
A failed check, a deploy that went wrong, or Sentry — all of them open the same incident, page the person holding the rota, and put “investigating” in front of customers before the first ticket lands.
Built for teams who still answer “is it down?” in Slack
The right person, not the whole channel
A rota decides who is holding the pager, layers cover nights and weekends with somebody else, and the escalation keeps going until one of them acknowledges. An alert at 3am does not die in a muted channel.
Severity decides what happens
One scale, P1 to P4, chosen once per component. A P1 on Billing can wake the on-call and go red on the status page while a P4 waits for the morning and never leaves the building.
Everything opens the same incident
Our checks, Sentry, Grafana, your CI, or two lines of curl from a deploy script. Close it from a different system than opened it — you agree on a name, not on an id you have to store.
FAQ
- Is the free plan really free forever?
- Yes for the whole loop: monitoring, on-call rotas, escalation with acknowledge, incidents, maintenance, and a status page on your own domain. Unlimited people — we do not charge per seat and do not intend to. What costs money is what costs us money: phone calls and SMS, and integrations we have to keep working against somebody else’s API.
- Do you page me, or just show a dashboard?
- We page you, and we keep going. Steps fire in order until somebody acknowledges, so nothing rests on one person seeing one message. Free channels are Slack, Telegram, webhooks, browser push and email; phone calls and SMS are the paid ones, because each one has a bill attached.
- Who decides where my number goes?
- You do, and only you. An admin cannot type a colleague’s phone or Telegram in on their behalf — each person confirms their own, by opening a link we email or messaging our bot. Until they confirm, nothing is sent there, and the rota shows it as unconfirmed rather than pretending it works.
- What can you actually watch?
- A URL, with a status code and optional text that has to be on the page — an error page rendered with 200 passes every check that only reads the code. A TCP port, for a database or queue with no health URL. A TLS certificate, which starts failing while there is still time to renew rather than on the day. And a heartbeat, for a backup or cron job whose failure is silence.
- Can other tools open incidents here?
- Yes. Pick the tool from the catalogue, choose which component it reports on, and paste the URL into it. Sentry, Grafana, Alertmanager, Datadog, CloudWatch, GitHub Actions, GitLab CI, UptimeRobot, Better Stack, Zapier — all free, because they call a URL and there is nothing for us to keep running. Your own scripts use the same API with a token.
- Do I have to store an id to close an incident?
- No, and you should not have to: the system that opens an incident is often not the one that closes it. You choose a name for the problem, send it when you open, and send the same name to close. Sending it twice does not open a second incident, so a retrying job stays harmless.
- Can I use a custom domain?
- Yes, on the free plan. Add status.yourapp.com in the page setup, point a CNAME and a TXT record at us, and hit verify. Removing the “Powered by” line is the paid part.
- Why not UptimeRobot or Better Stack?
- UptimeRobot gives you far more monitors — 50 on the free plan — but it caps you at five integrations, Slack sits on the paid tiers, and a status page on your own domain is paid. Better Stack is the closest match and is genuinely good: 10 monitors, a status page on your domain, all free. The difference that matters is on-call — its escalation and acknowledge start at $29 per responder per month, and here they are the free plan. Checked August 2026.
Coming from a paid tool? Read IncidentFlare vs Atlassian Statuspage or vs PagerDuty for 1–3 person teams.